Gaze tracker and method for tracking gaze thereof

1. A gaze tracker comprising:

an image acquirer configured to acquire eye images of a user using at least two cameras and at least two lightings; and

a processor configured to calculate a gaze to detect a lighting reflection point from each of the eye images and to extract a cornea central point, the processor configured to detect a pupil central point from each of the eye images and calculate a gaze vector using the cornea central point and the pupil central point,

wherein, when at least two lighting reflection points are detected from each of the eye images, the processor is configured to calculate vectors for connection between each lighting and the lighting reflection point corresponding thereto and to detect an intersection point between the calculated vectors as the cornea central point, and when only one lighting reflection point is detected from one or more eye images among the eye images, the processor is configured to extract the cornea central point using distance information between a pre-stored lighting reflection point and the cornea central point and the one lighting reflection point.

2. The gaze tracker according to claim 1 , wherein the processor is configured to calculate distances from the cornea central point to the at least two lighting reflection points and to accumulate and store calculated distance information.

3. The gaze tracker according to claim 1 , wherein the processor is configured to convert 2D coordinates of the lighting reflection point and the pupil central point into 3D coordinates.

4. The gaze tracker according to claim 1 , wherein the processor is configured to determine a lighting that forms the one lighting reflection point at a cornea among the at least two lightings using a face direction vector.

5. The gaze tracker according to claim 4 , wherein the processor is configured to calculate a vector for connection between the one lighting reflection point and the determined lighting and extract the cornea central point using the calculated vector and the distance information.

6. The gaze tracker according to claim 1 , wherein the processor is configured to calculate similarity between lighting reflection points in an eye image from which the at least two lighting reflection points are detected and an eye image from which one lighting reflection point is detected among the eye images, and to determine a lighting that forms the one lighting reflection point at a cornea.

7. The gaze tracker according to claim 1 , wherein the processor is configured to obtain the distance information by calculating an average distance of distance information for a predetermined recent period among distance information items from a pre-stored cornea central point to a lighting reflection point.

8. A method for tracking gaze of a gaze tracker comprising at least two cameras and at least two lightings, the method comprising:

acquiring eye images of a user using the at least two cameras;

detecting a lighting reflection point from each of the eye of the images;

upon detecting at least two lighting reflection points from each of the eye images, calculating vectors for connection between each lighting and a lighting reflection point corresponding thereto and detecting an intersection point between the calculated vectors as a cornea central point;

upon only one lighting reflection point from one or more eye images among the eye images, detecting the cornea central point using distance information between a pre-stored lighting reflection point and the cornea central point and the one lighting reflection point;

upon detecting the cornea central point, extracting a pupil central point from the eye image; and

calculating a gaze vector using the cornea central point and the pupil central point.

9. The method according to claim 8 , wherein the detecting of the lighting reflection point includes converting 2D coordinates of the lighting reflection point into 3D coordinates.

10. The method according to claim 8 , wherein the detecting of the intersection point of the vectors as the cornea central point includes calculating distances from the cornea central point to the at least two lighting reflection points and accumulating and storing calculated distance information.

11. The method according to claim 8 , wherein the detecting of the cornea central point using the one lighting reflection point comprises:

determining a lighting corresponding to the one lighting reflection point;

calculating a first vector for connection between the one lighting reflection point and the determined lighting; and

extracting the cornea central point using the first vector and the distance information.

12. The method according to claim 11 , wherein the determining of the lighting corresponding to the one lighting reflection point comprises determining a lighting that forms the one lighting reflection point at a cornea among the at least two lightings using a face direction vector.

13. The method according to claim 11 , wherein the determining of the lighting corresponding to the one lighting reflection point comprises calculating similarity between lighting reflection points in an eye image from which two or more lighting reflection points are detected and an eye image from which one lighting reflection point is detected and determining a lighting that forms the one lighting reflection point at a cornea.

14. The method according to claim 8 , wherein the distance information is obtained by calculating an average distance of distance information for a predetermined recent period among distance information items from a pre-stored cornea central point to a lighting reflection point.